Open Source

A free, open-source macOS menu-bar app that reads what Claude Code and Codex are thinking while they work. Write any rule in plain English: skipping tests, touching secrets, working around a blocker, claiming done without proof. ThoughtCop checks every thought and reply against your rules and flashes a red bolt the moment one is broken. No hook, no proxy, nothing changes in the agent.

MergeKit Cloud helps developers work with open-weight LLMs through model merging, benchmarking, and capability comparisons. Beyond merging models, it helps developers measure performance changes, identify trade-offs, and evaluate results across reasoning, math, and coding tasks. Manage experiments, organize recipes, and explore deployment workflows in one platform. Built for AI developers, researchers, and startups.

I built ej for a narrower problem than general classification: given some state and a set of decision questions whose options can be defined at runtime, return a probability distribution for each question in one pass. It handles choice, yes/no, and ordered scores, runs on CPU, needs no token generation, and ships as one 11.4 MB model file. The repo includes training, evaluation, adaptation, packed weights, and the benchmark harness.

slash-editor gives your React app the Notion block experience: a slash menu, drag handles, nested reorder, turn-into, comments and real-time collab. It runs on a headless Tiptap v3 core. The UI arrives through a shadcn registry, so the components land in your repo and follow your Tailwind tokens. Tables, columns, Mermaid, AI via your own model, markdown import/export. MIT end to end, no paid tier, nothing hosted.

WeftCut is a free, open-source desktop video editor your AI agent can control through MCP. Connect your agent to trim clips, add captions, build animations from code, and edit a real timeline while you watch. Review changes, undo, or take over by hand. Available for Windows, macOS, and Linux.

CodeAF is a coding harness built for open models, to get the most out of every dollar. Frontier-grade coding on open models, at a fraction of the cost.I t is also a different way to work once more than one thing is going on: instead of three terminals of agents with you in the middle, one window where you hand work off, see what is moving across every project, and step in only where your judgment is needed.

Everything Grok Bot, Muse and Dots do, open source and 100% free. Run Claude Code, Codex, Cursor or 40+ agents as named bots on your own computer, then reply and approve from your iPhone, Mac, Linux or terminal. MIT, end-to-end encrypted.

JarvisCore is a runtime where AI agents operate as a fleet of equal peers. Agents discover one another by capability, execute tasks from a shared ledger, and authenticate to every external service through a zero-trust broker, never with their own keys.

Arc is a library of React components and blocks with motion built in. The core is free and open source (MIT) and installs with the shadcn CLI, so the code lives in your project and you can change anything. Every component has Markdown docs, llms.txt and an MCP server, so Claude Code or Cursor can pick and install the right piece instead of inventing one. Pro adds showpiece motion components and full app screens. New components ship every day.

When you run Claude Code, Codex CLI or OpenCode in several folders at once, each session stops now and then to ask for permission, and you only find out when you look at its terminal. OpenCompanion starts those sessions, lists them on one screen, tells you when one is waiting, and lets you answer from the desktop or from your phone on the same Wi-Fi. No account, no cloud server, no telemetry. Open source (MIT), for Windows, Linux and macOS.

For years we've used LLMs built to write text to make choices. JevGPT returns the favor: a chat app where TypeSafe's Jev, a model built to make choices, writes every reply one multiple-choice word at a time. Does it work? Sort of. Open source.

Run coding agents like Claude Code and Codex on infrastructure you control and manage them from desktop or mobile. It's open source, self-hostable, and built in Rust.

A status bar for any terminal.

Kolibri is Europe's newest open-source AI model by Aleph Alpha. An English-German Mixture-of-Experts model with 78B total and 3B active parameters, a context of up to 1M tokens and open weights under Apache 2.0, specialized for sovereign, mission-critical work. You can try out the model for free via AI providers like xprivo and Tesseracted. Kolibri beats Qwen2 635B, Nemotron 3 Super and Mistral Small 4.

devpit is a native desktop app for running Claude Code agents. One terminal you actually watch, a board where moving a card runs real work, a cost on every agent call, and an orchestrator that sees every project at once. Local-first, open source.

WriteCode is a browser-based coding workspace built for students and developers. Write, run and test code without setting up a local environment, with tools that make learning and debugging easier. It brings coding, execution, debugging and code visualization into one place, so you can go from writing your first line to understanding what your program is doing without switching between multiple tools.

Bentomux is a calm desktop app for running coding agents side by side: project workspaces, persistent split-pane terminals, live agent state and approvals.

Ship AI agents to production in hours. Opengeni gives you sessions that recover from failures, isolated sandboxes, 100+ integrations, human approvals, and visibility into every step and dollar spent. Focus on your agents. We handle the infrastructure. Self-host for free, or get started on our cloud.

A React component library of animated AI agent status orbs. Thinking, searching, compacting and more, built with attention to detail on every state and variant.

A desktop client for SSH, SFTP, Docker over SSH and Remote Desktop, on Windows and Linux. - GitHub - juddisjudd/bawkterm: A desktop client for SSH, SFTP, Docker over SSH and Remote Desktop, on Windows and Linux.